The ontogeny of the human globin genes is an important focus of study both with respect to the fundamental developmental biology and molecular genetics of the control of expression of this complex gene system, but also because modifying this developmental control would be of therapeutic value in the treatment of the prevalent genetic diseases of hemoglobin, such as sickle cell anemia and thalassemia. We are studying this problem from several aspects. First, we have improved culture methods for human erythroid precursors, by antibody selection, so as to obtain purified populations of these cells to establish at which developmental stage drugs change their phenotype. Second, we have developed real-time, quantitative PCR methods to measure globin gene expression in single cells and have used these methods to analyze the mechanisms of the effects of chemically useful drugs, such as hydroxyurea and butyrate, on globin gene expression. Third, we have shown that nitric oxide donors, such as cysteine-NO, can induce fetal hemoglobin expression in K562 cells as well as the purified erythroid (CD34) precursor cells described above. Induction of fetal hemoglobin was demonstrated both by quantitative PCR of gamma-globin mRNA and by HPLC of fetal hemoglobin protein. Further, we have shown that both NO-donors and hydroxyurea acts by stimulating guanylyl cyclase and increasing cyclic GMP levels;conversely, inhibitors of guanylyl cyclase block the action of both agents. Agents to increase fetal hemoglobin by this mechanism are being developed in several laboratories and being readied for clinical testing. In more recent studies we have shown that hydroxyurea induces NO production in macrophages and endothelial cells-in part through inhibition of proteasome degradation of the eNOS protein-and that this efect may allow a mechanism for a paracrine efect of NO on erythroid progenitors from these signalling molecules produceed by stromal and other bone marrow cells. In current studies of globin production-at the mRNA and protein levels-we are studying the mechansims of such singaling in hemoglobin ontogeny and its therapeutic implications. Hydroxyurea, a drug widely used for treating myeloproliferative diseases, has also been approved for the treatment of sickle cell disease by raising fetal hemoglobin (HbF). We demonstrate now that during erythroid differentiation, endothelial NO synthase mRNA and protein levels decline steadily, as does the production of NO derivatives and cAMP levels, but cGMP levels are stable. Hydroxyurea increased intracellular cGMP levels and cAMP levels in erythroid progenitor cells (EPC). The NO donor, DEANONOate, induced much higher cGMP levels, but reduced cAMP levels. Hydroxyurea (1 mM) induced production of approximately 45 fmol cGMP/min/ng of purified sGC, similar to induction by 1 ?M of DEANONOate. We found that hydroxyurea and ProliNONOate produced iron-nitrosyl derivatives of sGC. Thus, we confirm that hydroxyurea can directly interact with the deoxy-heme of sGC, presumably by a free-radical nitroxide pathway, and activate cGMP production. These data add to an expanding appreciation of the role of hydroxyurea as an inducer of the NO/cGMP pathway in EPC. These mechanisms may also be involved in the cytostatic effects of hydroxyurea, as well as the induction of HbF. Overall, these results raise the possiblity that agents which affect this cyclic nucleotide pathway, or other signaling pathways, can be used to increase fetal hemoglobin levels in patients with sickle cell anemia and thalassemia. Such therapeutic advances have been the focus of this research program for almost three decades. In other recent studies we have been analyzing the change in fetal hemoglobin levels in normal newborns with age and find-using both quantitative protein and mRNA measurements-that there appear to be two distinct silencing mechansims for fetal hemoglobin-one primarily cellular and one primarily with respect to transcription mechanisms. We are now studying these processes in sickle cell children and will examine the effects of drugs on these silencing mechanisms. Further analyses of erythroid cells in culture are underway with respect to gene expression patterns in these cells and several manuscripts on this work have recently been submitted or published.